printf/sprintf take a list of arguments. In the first case sprintf gobbles up all the remaining items in the hash initialization list (even though the format string doesn't require them). In the second case the list that sprintf sees is constrained by the parens.

Note that => and , (comma) are in essence the same except that => stringizes the item to the left of it.

Perfect. Thanks++

Where do you want *them* to go today?

Deparse is your friend (though in this case, it needs a little perltidy help to be readable):

$ perl -MO=Deparse,-p|perltidy
  my $item = { 
    JOBN           => $record->[0],
    CUSTOMERN      => $record->[1],
    JOBDESCRIPTION => $record->[2],
    QUANTITY       => $record->[3],
    QUOTEDSALES    => sprintf "%0.2f", $record->[4] * 0.04,
    DUEDATE        => $record->[5],
    PRODPLANNER    => '5',
    PRODUCTCODE    => '900',
    SALESMANN      => '48',
  };
- syntax OK
(
    my $item = {
        'JOBN',
        $$record[0],
        'CUSTOMERN',
        $$record[1],
        'JOBDESCRIPTION',
        $$record[2],
        'QUANTITY',
        $$record[3],
        'QUOTEDSALES',
        sprintf( '%0.2f',
            ( $$record[4] * 0.04 ), 'DUEDATE',   $$record[5],
            'PRODPLANNER',          '5',         'PRODUCTCODE',
            '900',                  'SALESMANN', '48' )
    }
);
[download]

The commas in the hash are pair separator. After the sprintf they become argument separator for sprintf, and without the parens it lasts until the last scalar (48 in this case). With parens, the commas as argument separator for sprintf are only before the closing paren. The closing paren itself acts as argument terminator for sprintf, separating it as an execution unit from the rest of the pairs. After the closing paren, commas become pair separator again.